The moral case for human music

If you listen closely, you can hear the slop. An Apple Music executive said in April that more than one-third of the tracks being delivered to that streaming service were “100 per cent AI”. According to the French streaming platform Deezer, 90,000 fully AI-generated tracks per day were uploaded there in June, more than quadruple the number a year before. The software responsible for a substantial proportion of this deluge is Suno, the leading AI music-generation app. In November last year, the company reported that more than seven million songs were being generated on the platform every day.

In a January 2025 appearance on The Twenty Minute VC, a venture-capital podcast, Suno CEO and co-founder Mikey Shulman said that his company’s mission is to make the joy of music-making accessible to billions of people. A clip from this interview went viral among musicians — not because they applauded his sentiments, but because it enraged them.

“It’s not really enjoyable to make music now,” Shulman said. When asked to elaborate, he said, “It takes a lot of time, it takes a lot of practice, you have to get really good at an instrument or really good at a piece of production software. I think the majority of people don’t enjoy the majority of the time they spend making music.” The implication appeared to be that Suno offers customers the experience of making music without any of the tedium or expense involved in doing it the old-fashioned way. What could be better than that?

Reflecting on this clip, and the degree to which it incenses musicians, reveals a strange gap in the way AI music is talked about. Public discussion of musicians’ grievances against AI tends to focus on the threat that AI music generation poses to their livelihoods — lost streaming royalties or fewer commissions from film and TV — and the widespread copyright infringement that the training of music AI models appears to have involved. These material injuries are real and weighty. However, the musicians infuriated by Shulman’s comments don’t appear to be responding to the threats posed to their livelihoods. What is their grievance exactly, and is it legitimate?

The worry can’t be that AI-generated music must sound terrible: it’s getting to the point where many people can’t tell if a piece of music is AI-generated. And it obviously can’t be that musicians object to the use of fancy digital tools to make music; they use these all the time and are happy about it. The real grievance is moral in nature — and Shulman’s comments help us to diagnose what exactly this grievance is. Specifically: promoting an AI music-generation app as a way of making music requires promoting the view that making music is a mere means to an end, with no value in its own right. This view is false. And since musicians are people who have built their lives around the value of making music, the promotion of AI music-generation software constitutes a public insult to musicians — one that is built on a misunderstanding of what it is to make music, and why it matters.

Making music is hard, especially when you’re a beginner. But does that fact justify using AI to do it? In general, hardship is a reason to skip an activity only if that activity doesn’t have some other value that compensates for the hardship. For example, the effort involved in hiking to the top of a mountain is a reason to take the cable car only if hiking to the top doesn’t have a value that compensates for the effort. And so, using AI to make music only makes sense if music-making has no value that compensates for the hardship involved in learning to do it.

Thus, the promotion of AI music-generation software tacitly construes music-making as what I’ll call a “purely telic” activity. (I’m drawing on the MIT philosopher Kieran Setiya’s discussion of “telic” activities here, telos as in “end” or “goal”, in his excellent book Midlife.) A purely telic activity is one directed at a terminal state of having done something (the “telic” part), where the activity has no value beyond that of achieving the terminal state (the “purely” part). Construed as a purely telic activity, music-making has no value over and above that of having produced a piece of music, and so the more efficiently you can get to that point, the better. AI music generators make perfect sense on this view. The problem, of course, is that the purely telic view of music-making is false. For a musician, the point of making music isn’t merely to end up having made a piece of music: the actual making is part of the point.

That’s not to say that making music is pleasurable all the time. It often involves moments of friction: the high note you’re not hitting, the rhythmic “lock” that just isn’t happening between you and the drummer. But when you push through those moments of friction and get to the other side, what awaits can be some of the most exhilarating, transcendent experiences it’s possible to have as a human being. Those moments more than compensate for the hardship that the friction brings. In fact, they require the friction. Eliminate the friction, and you eliminate the thrill of overcoming it.

The purely telic view of music-making is false from the perspective of the listener, too. Why do people shell out small fortunes to see their favourite musicians play live? It’s not just because they want to hear their “musical outputs”. They can do that at home, by playing their recordings. Nor is it just because they want to hear the music in the company of other fans, while simultaneously seeing the musicians. Imagine that you’ve forked out $300 to see your favourite band. Everyone files into the stadium, fizzing with anticipation. The lights go down, the band comes out — and then, someone presses “play” on a laptop, the album starts playing on the house speakers and the band stands there waving at the audience for the rest of the evening. You would be outraged. Because what you wanted, what you paid for, was to experience the band making their music. You wanted to witness these people actually bringing their music into being, with their bodies, deploying the skill and understanding that is required to do this.

Our interest in music-making as a spectacle in its own right distinguishes music from other arts that are threatened by generative AI. While it matters to us that novels, say, be written by human novelists, we don’t have the same interest in the process of novel writing as we do in the process of music-making. If someone were to spend money to watch their favourite novelist write, we would probably wonder why on earth they had done that.

In a world where AI-generated music dominated, it’s not clear that there would be any room for our interest in music-making as spectacle. The process by which AI music generators produce music is opaque. Whatever it is that happens between prompt and output, nobody – not the prompter, not the software designers, certainly not you or me – can witness it. One could watch AI music “being made” in the sense of watching someone prompt the software. But it’s hard to imagine throngs of fans gathering to watch someone hunched in front of a chatbot window, typing things like, “Make a jazz song about watering my plants.”

An AI music exec might object along the following lines: “OK, so making music with our app looks pretty different to making music the old-fashioned way. But we’re just giving ordinary folk a more accessible way of making music. And besides, our software can be used to help musicians make music too: it can be used as a powerful editing tool, for example. We’re not trying to replace musicians! We love musicians!”

There are a few things to say in response. First, AI music apps are not giving ordinary people a more accessible way of making music. They are giving them an opportunity to do something that looks like “making music” only if you adopt the purely telic view of music-making — which, as I’ve argued, is false. If you’re someone who wants to experience the joy of making music, you want to do the thing that musicians do, the thing that humans have valued for the past 35,000 years or so. That is: you want not merely to have made music, but to make it. AI music execs speak as though tweaking prompts and evaluating outputs on an app is a way of making music — but it’s not. It’s a way of commissioning and critiquing music. There are many, many ways of making music — look around the world and you will see thousands — but entering prompts into an app isn’t one of them.

Second, there are far cheaper ways to bring the joy of music-making to the masses: for example, funding music education in schools, buying instruments for low-income families or paying musicians (what a thought) to give music lessons. None of these would require putting the world’s music into a blender so that AI models can spit out sound without acknowledging the musicians on whose creative labour those models rely.

Third, it’s true that if they’re used in specific ways, AI music apps might be useful to musicians, as Bob Moog’s synthesisers turned out to be. But that doesn’t mean that an AI music app is the same sort of thing as a Moog. Moogs were built and marketed as technology that offered musicians new ways of making music. A Moog doesn’t make music for you: it just gives you more options as to what and how to play. That’s why musicians love them. But when an AI music app is used to do the thing it was built to do, it doesn’t help anyone to make music: it makes it for them.

Fourth, music AI firms are not music-technology firms. A music-technology company makes products for musicians. AI music companies make products primarily for two other groups: people who want to experience what it’s like to make music and can be persuaded they can do so without putting in any effort, and people who need music for commercial reasons and don’t care how it’s made as long as it’s serviceable and cheap. Neither group consists mainly of musicians, and in neither case is the product helping anyone experience what it’s actually like to make music.

Finally, individual AI developers may indeed respect musicians. But it’s in music AI firms’ economic interest to encourage the view that human music-making can be replaced with something else, with no loss of value. The more people believe that, the greater the uptake of AI music-generation apps will be. Individual music AI developers may not think they’re trying to replace human music-making. But the logic of the enterprise suggests otherwise.

“We built Suno because we believe that making music is joyful, fun and deeply rewarding, and that everyone should get to experience it,” a company spokesperson said in a statement. “The majority of our company are musicians themselves and a deep respect for the craft of how songs get made is embedded into everything we do.” Referring to his comments about music-making, Shulman said in March, “I really wish I had chosen different words,” and “I do have a lot of respect for music.”

Recommended

It is possible to attempt to persuade people that an activity has no value in its own right without insulting those who perform that activity. For instance, if I were to try to convince everyone that doing laundry is valueless, those who do laundry wouldn’t be insulted. But the music case is different. For many musicians, making music isn’t just something they happen to do. They have built their lives around the value of making music: they have made real, palpable sacrifices for the sake of that endeavour that have had a profound impact on the shape of their lives. And so, the promotion of the view that human music-making has no value in itself amounts to a public insult: an expression of disrespect not just for what musicians do, but for who they are. And insofar as that message is intended to persuade, it invites others to share that disrespect.

Music AI firms pose real material threats to musicians. But when musicians rail against them, they are responding to a moral injury as well as a material one. This moral injury is just as real, and it would remain even if every material threat were addressed. AI boosters may say that their hostility is just fear of change. But they’re wrong. The promotion of AI music-generation software as a way of making music is an insult to musicians. Musicians are right to react to it as such.

Jenny Judge is a senior lecturer in philosophy at the University of Melbourne and half of the electro-acoustic duo Pet Beast

Find out about our latest stories first — follow FT Weekend Magazine on X and FT Weekend on Instagram